| Author |
Message |
firsttimedesigning
Joined: 04 Jul 2007
Posts: 97
|
 17 Feb 2008 9:38 Full Differential and Single Ended |
|
|
|
| I can understand why full differential circuits are better than the single ended ones but what I dont understand is how do you implement a fully differential circuit. Take comparator for example, all the comparators that I see in the book have single ended output. None of them has double ended outputs. So how do I actually make it a "fully differential" comparator?
|
|
| Back to top |
|
 |
strabush
Joined: 10 Apr 2002
Posts: 83
Helped: 6
|
| 17 Feb 2008 10:38 Re: Full Differential and Single Ended |
|
|
|
You can take a differential in and out OpAmp, remove its compensation capacitors (as no feedback will be used), and use it as a comparator.
If open loop of the OpAmp is high, it will function well as comparator.
|
|
| Back to top |
|
|
FvM
Joined: 22 Jan 2008
Posts: 2476
Helped: 408
Location: Bochum, Germany
|
| 17 Feb 2008 11:21 Re: Full Differential and Single Ended |
|
|
|
Hello,
very fast comparators (Anlog Devices ADCMPxx et al) typically have differential outputs, cause they should be able to drive differential logic, e. g. LVDS. Although at the digital side the motivation for differential signaling isn't exactly the same as for the analog side, it shows similar advantages, e. g. lower interference to adjacent signals.
Regards,
Frank
|
|
| Back to top |
|
|
firsttimedesigning
Joined: 04 Jul 2007
Posts: 97
|
| 19 Feb 2008 10:22 Re: Full Differential and Single Ended |
|
|
|
Ok thank you for replying. but i am still not sure on how to connect the op-amp. I have drawn a picture but i am not sure if i have connected it correctly. Also, I still dont understand, how is the op-amp(double ended output) that i have drawn better than the single ended one?
|
|
| Back to top |
|
|
FvM
Joined: 22 Jan 2008
Posts: 2476
Helped: 408
Location: Bochum, Germany
|
| 19 Feb 2008 10:30 Re: Full Differential and Single Ended |
|
|
|
Hello,
shorting the differential op output to ground is no good idea, it would cause excessive output current and probably damage the device. However I don't see, why you give an integrator circuit but asked for a differential comparator. Integrator circuit would give no meaningful results with a comparator. So I wonder what you actually try to achieve?
Regards,
Frank
|
|
| Back to top |
|
|
firsttimedesigning
Joined: 04 Jul 2007
Posts: 97
|
| 20 Feb 2008 5:05 Re: Full Differential and Single Ended |
|
|
|
Thanks for the reply.
Ok yeah so actually I am having trouble with both circuits, integrator and comparator. What I would like to achieve is that I want to connect the + end of the integrator to a comparator and then connect the output of the comparator to a control logic. I have drawn a new circuit, please tell me if it is right. Instead of shorting the - output to the ground, i just open it.
|
|
| Back to top |
|
|
FvM
Joined: 22 Jan 2008
Posts: 2476
Helped: 408
Location: Bochum, Germany
|
| 20 Feb 2008 10:23 Re: Full Differential and Single Ended |
|
|
|
Your circuit should work, but isn't using the fully differential feature anyhow. Why not extending it to true differential circuit?
|
|
| Back to top |
|
|
firsttimedesigning
Joined: 04 Jul 2007
Posts: 97
|
| 21 Feb 2008 3:36 Re: Full Differential and Single Ended |
|
|
|
Thank you for the replying but it seems to me that if i make it true fully differential circuit, then the size of the circuit will increase. the capacitor, for example, i will need two of them. is there anyway to make it true fully differential without adding another cap and res. I guess not...
I have drawn a picture of the output of the integrator and comparator. not sure if they are correct. and so far I have not see the advatange of making the ADC differential. The -end output just seems like it can be eliminated...i dont get it, are all ADCs suppose to be differential? if so, why? i read papers and it seems to me that the ADCs presented are differential...but -end output, how do i use it to make the ADC better? i am really confused.....
|
|
| Back to top |
|
|
nicolas.borfigat
Joined: 02 Mar 2008
Posts: 12
|
| 02 Mar 2008 15:48 Full Differential and Single Ended |
|
|
|
Hello,
Maybe I make a mistake but one of the main reason for using a fully diff device is it ability to cut down the noise.
|
|
| Back to top |
|
|
FvM
Joined: 22 Jan 2008
Posts: 2476
Helped: 408
Location: Bochum, Germany
|
| 02 Mar 2008 16:03 Re: Full Differential and Single Ended |
|
|
|
| Yes, when noise is understood as common mode interferences, ground bounce and such. Additionally, linearity is improved, even harmonics are supressed, emission of interferences is reduced.
|
|
| Back to top |
|
|
firsttimedesigning
Joined: 04 Jul 2007
Posts: 97
|
| 04 Mar 2008 9:14 Re: Full Differential and Single Ended |
|
|
|
Fvm
Actually, I just did a hand analysis and I found out that the circuit you have posted would not integrate at all. but thx for replying anyway...
|
|
| Back to top |
|
|
FvM
Joined: 22 Jan 2008
Posts: 2476
Helped: 408
Location: Bochum, Germany
|
| 04 Mar 2008 9:36 Re: Full Differential and Single Ended |
|
|
|
| Quote: |
| I just did a hand analysis and I found out that the circuit you have posted would not integrate at all. |
Do you know a way, to convince the hardware versions of the circuit, that are performing well since years, they are simply doing wrong?
|
|
| Back to top |
|
|
firsttimedesigning
Joined: 04 Jul 2007
Posts: 97
|
| 04 Mar 2008 10:11 Re: Full Differential and Single Ended |
|
|
|
| thx for the reply, but if you can, show me your hand analysis and prove me wrong. cuz i dont think i made a mistake...and i believe spice didnt either...
|
|
| Back to top |
|
|
wateror
Joined: 26 Aug 2007
Posts: 61
|
| 13 Mar 2008 3:26 Full Differential and Single Ended |
|
|
|
| refer to allen's book... you put a inverter at each of the opamp output, then you get differential outputs.
|
|
| Back to top |
|
|
kivvzhou
Joined: 09 Aug 2006
Posts: 22
Helped: 2
|
| 20 Mar 2008 7:52 Full Differential and Single Ended |
|
|
|
| for diff opamp,guay's book explain very good!
|
|
| Back to top |
|
|
jecyhale
Joined: 19 Feb 2008
Posts: 431
Helped: 46
Location: China
|
| 26 Mar 2008 11:06 Re: Full Differential and Single Ended |
|
|
|
| firsttimedesigning wrote: |
| I can understand why full differential circuits are better than the single ended ones but what I dont understand is how do you implement a fully differential circuit. Take comparator for example, all the comparators that I see in the book have single ended output. None of them has double ended outputs. So how do I actually make it a "fully differential" comparator? |
Differential signal is to redcue the noise, so it is good for small signals not for large siganl. comparator output is large signal.
|
|
| Back to top |
|
|
FvM
Joined: 22 Jan 2008
Posts: 2476
Helped: 408
Location: Bochum, Germany
|
| 26 Mar 2008 12:08 Re: Full Differential and Single Ended |
|
|
|
| Really fast comparators usually have differential ECL or LVDS outputs. So actually differential signaling is suitable for digital signals or large signal as well.
|
|
| Back to top |
|
|
cretu
Joined: 12 Nov 2003
Posts: 138
Helped: 4
|
| 28 Mar 2008 8:20 Re: Full Differential and Single Ended |
|
|
|
| FvM wrote: |
| Really fast comparators usually have differential ECL or LVDS outputs. So actually differential signaling is suitable for digital signals or large signal as well. |
It would also help if you don't realy need rail to rail signal through your high speed chain. You can have only at the output stage something that will be rail-to-rail.
|
|
| Back to top |
|
|
BackerShu
Joined: 22 Dec 2007
Posts: 23
|
| 31 Mar 2008 12:35 Re: Full Differential and Single Ended |
|
|
|
| kivvzhou wrote: |
| for diff opamp,guay's book explain very good! |
Sorry,who is guay?
Do you mean gray?
|
|
| Back to top |
|
|
MicroElec
Joined: 31 Mar 2008
Posts: 3
|
| 31 Mar 2008 17:28 Full Differential and Single Ended |
|
|
|
Hello,
A Fully Differential comparator is a comparator with 2 inputs and 2 outputs....the aim is to reduce the Ground noise.
To make a Fully Differential comparator you can take for exemple un dif stage(with 4 transistors + current source).
The outputs are taken between each half part in the circuit (drains).
hope that i helped you.
|
|
| Back to top |
|
|
northeast1
Joined: 21 Jun 2007
Posts: 150
Helped: 4
|
| 13 May 2008 9:50 Re: Full Differential and Single Ended |
|
|
|
| firsttimedesigning wrote: |
| I can understand why full differential circuits are better than the single ended ones but what I dont understand is how do you implement a fully differential circuit. Take comparator for example, all the comparators that I see in the book have single ended output. None of them has double ended outputs. So how do I actually make it a "fully differential" comparator? |
I think it is better for small signal, not large signal process.
|
|
| Back to top |
|
|
zlatkoMM
Joined: 04 Apr 2008
Posts: 138
Helped: 5
Location: Croatia
|
| 13 May 2008 14:21 Re: Full Differential and Single Ended |
|
|
|
| Full diferential is just story!
|
|
| Back to top |
|
|
